No abstract

No abstract

No abstract

No abstract

Detecting pedestrians using conventional optical camera has got many problems. It's difficult to be used to detect pedestrian using only single optical camera. Detecting pedestrians in a crowded environment from other objects is a very difficult task. This paper presents a method to detect multiple pedestrians from a moving camera. The detection component involves a cascade of modules. We used a supervised self organization map neural networks as our classification mechanism. First, each frame is divided into four parts then our proposed fast BMA (Block Matching Algorithm) is used to obtain four representative motion vectors from two consecutive input frames. Then frame differencing method, based on obtained representative motion vectors is applied to generate differenced image. Second, pedestrians are detected by the step that the differenced image is transformed into binary image, two level of noise reduction is then applied and then we used artificial neural networks as a second level of classification.